Re: Position of ALT text on IMG

On Mon, 08 Nov 1999 06:08:03 -0800, you wrote:

>>>There is the general issue of placeholder content of which the ALT attribute
>>>on the IMG tag is one example.
>> The ALT attribute for IMG has _one_ single purpose only, to be available
>> as "alternative content" presented _inline_ instead of the graphic that
>> for one reason or another is not shown.
>As you say, "instead of a graphic, *for one reason or another*", your
>statement itself conveys the multiple uses or at least conditions under
>which the ALT attribute is used.

So let me rephrase "instead" then, and quote the specs...

RFC1866 states the following for the ALT attribute of an IMG element...

  "HTML user agents may process the value of the ALT attribute as an
   alternative to processing the image resource indicated by the SRC

...and as a definition of the ALT attribute...

     text to use in place of the referenced image resource,
     for example due to processing constraints or user

and has earlier in the spec also stated this...

  "The `NAMELEN' parameter in the SGML declaration (9.5, "SGML
   Declaration for HTML") limits the length of an attribute value to
   1024 characters."

...there should be a lot of interesting _ALTernative_ text that could be
given within 1024 characters I guess.

For HTML4, instead of quoting it, I'l point to section...

  13.8 How to specify alternate text

...that gives another bunch of good info on what the ALT attribute is to
be used for.

And for HTML4 'NAMELEN' has been increased to 65536 characters.

>My point was only that one such reason/condition is while an image is being
>downloaded / is incomplete.

But _if_ image download is activated the ALTernative text should _not_
be rendered, so what's the problem?


>You did get the invite for the MacIE5 beta, yes?

I don't have a Mac, have never had a Mac, and as soon as there's a
decent CSS capable non-MS browser _released_ for a *NIX environment I
will kick MS-WinNT out the door and never touch an MS product again, in
my private life at least. (nothing personal, it's just the way I feel
about MS after having followed them and used their products since 1982)

[yea, I know there are Win emulators for *NIX, so I could run Opera, but
that's not what I'm thinking of]


>> Also, the idea of a browser that "when I turn graphics off" still
>> insists on showing "placeholders" for those graphical pieces is just
>> absurd.

>If you are viewing with a graphical browser, the reason is quite
>straightforward.  Many pages depend on the boxes for images being laid out
>for the proper layout of the entire page.  If you take away those boxes, but
>still render, say the margin, borders, padding, float, width, height etc. of
>everything else, you will typically have a completely different layout than
>the author intended.

So what? It's not the author who decides what I want to do with a web
page. I may have reasons of my own to "process" its content in a ua the
way I want it and at such times I might not care at all about "layouts".
Web publishing _is_not_ about DTP in the first hand, and it's not about
letting publishers "put on a show" on users VDU's either, unless we
allow that to happen by controlling the ua to let it happen.

But I know, the whole idea of the www has changed over the years, it's
been slowly "taken over" by publishers that would much prefer to really
push their views on to users if they could.

>Now as for the dorky little "placeholder graphic" that are show in the box -
>I have to agree with you there - it would be nice to have a mechanism to not
>show the icon, and just show ALT text flowed into the box which was laid out
>for the image.

All 65536 characters? if it came to that? Would be very impractical, no?

>Those "placeholder graphics" are typically built into the UA - so no extra
>downloading is happening.

There are times when it may be a good idea _not_ to include the size of
an image in either the stylesheet or the markup. At such times it would
be difficult to come up with a "placeholder" that would not "ruin the
layout" of course, unless an IMG  download is done in the background to
find out about the IMG size anyway.

Jan Roland Eriksson <>

Received on Monday, 8 November 1999 18:06:16 UTC